In this study, we introduce a novel peptide sequence for cancer targeting, and further demonstrate its ability to enhance gene delivery to cancer cells. The receptor we are interested in targeting is the Urokinase-Type Plasminogen Activator Receptor (uPAR),
57 a class of molecules found to be over-expressed on cancer cell membranes, particularly those of the breast and prostate [171-172]. Otherwise termed as CD87, uPAR is a 313-residue, phosphatidylinositol (GPI)-anchored protein which has urokinase plasminogen activator (uPA) as its natural ligand for binding [107]. The cellular function of uPA has been described to act as a serine protease, digesting components of the extracellular matrix, by activating the zymogen plasminogen into its active protease state of plasmin [108]. Before activation, uPA is produced and stored in the cell as a single peptide chain termed ‘pro-uPA’, which is converted into its active form of uPA upon a single catalytic event. Once receptor-bound, the activity of uPA is inhibited as it complexes with two types of inhibitors, PAI-1 and PAI- 2. Internalisation by endocytosis is then only mediated on further complexation with the low density lipoprotein receptor-related proteins (LDLR), α2-macroglobulin receptor (α2MR) and with glycoprotein 330 (gp330) [109]. In regards to the mechanism of internalisation, it has been shown that the endocytosis of the LDLR- uPAR-uPA complex occurs via clathrin-coated pits, through which they enter into early endosomes altogether. The site responsible for the binding to α2MR was found to be on domain III of uPAR, a site which therefore cannot be blocked if endocytosis is to occur [176].
The structure of uPAR consists of 3 homologous domains, where the functional epitope for uPA-binding (Kd< 1nM) is located in the third loop of the N-terminal domain (Figure 1-16) [110]. Recently published by Haui et al. is the crystal structure of soluble uPAR in complexation with uPA [111]. At the resolution of 1.9Å, it was found that all three domains of the receptor participate in uPA-binding by forming a concave cone shape, into which uPA inserts. This cavity is described to have a width of 25Å and a depth of 14Å, factors which need consideration when designing uPA- like ligands for uPAR-binding.
58 Figure 1-16: Top, structure of GPI-anchored Urokinase Plasminogen Activator Receptor. Above representation shows the three domains and the amino acid sequence of the 55kDa receptor. Large arrows point to positions of phase-1 introns in the gene structure, and small arrows indicate regions sensitive to proteolysis and to GPI-specific phospholipase C (PI-PLC). The LU-domains (uPAR-specific domains I, II and III) in uPAR are labelled from the N- terminus by Roman numerals and the location of their putative loop structures are numbered in the model by Arabic numerals (1, 2 and 3). Adapted from Ploug et al 2002 [178]. Bottom, molecular surface representation of the ATF (or the GFD, growth factor domain) region of uPA (blue) fitting into the wide cavity formed by the three domains of the uPAR receptor (green, orange and purple). The binding cavity is found to have a width of 25Å and a depth of 14 Å. Adapted from Haui et al. Science, 2006 [111].
59 uPA on the other hand is a composite structure of three structurally independent regions: the amino-terminal growth factor domain, the kringle domain and the carboxy-terminal catalytic domain, the region in which the catalytic properties of uPA are exhibited [112-113]. The N-terminal region of uPA is termed the ‘amino-terminal growth factor domain’ due to its sequence homology with a part of the epidermal growth factor (EGF), found to be in the form of a single-chain peptide with three cysteine-bridged loops (Figure 1-17) [114]. The region responsible for binding was found to be located in the amino-terminal growth factor domain (residues 1-125), specifically in residues 12-32 (U21).
Figure 1-17: Crystal structures of ATF or growth factor domain (blue) and Kringle (grey) of domains of uPA. The bottom right corner of the ATF domain represents the Ω loop of the binding sequence of uPA. Derived from Haui et al. Science, 2006 [126].
1.7.1. Binding of uPA to uPAR
Peptide sequence of residues 12-32:
12 30 32
C D C L N G G T C V S N K Y F S N I H W C N C
It is within the amino acid sequence of 12-32 that studies by Appella et al. have shown the presence of two cyclic peptide structures that combine to make the binding sequence. The first loop is formed between amino acid residues of 13-20 LNGGTCV, termed the U7 sequence, and the second loop was found to be between residues 19-30, VSNKYFSNIHW, termed the U11 sequence. These studies further describe the binding affinity of the U11 sequence to be higher than that of the U7
60 sequence, although both do not surpass the binding affinity of the U21, which consist of both the U7 and the U11, suggesting that the presence of both the first and middle loops of the amino-terminal growth factor give increased binding to uPAR. However, using shorter peptide fragments as targeting ligands has benefits, such as ease of synthesis and easier control of secondary conformation. uPA fragments, instead of whole uPA as targeting ligands is considered advantageous, as the engineering process is greatly simplified, and the effects of the enzymatic properties of the C- terminal domain of uPA are avoided.
An example of the use of the U-7 peptide as a targeting ligand for uPAR-expressing cell demonstrates the incorporation of the synthetic peptide onto the surface of adenovirus vectors to retarget airway epithelial cells, which are also shown to express uPAR molecules [115]. Exposure of human airway epithelial cells to solutions of uPA and U-7 showed to stimulate the endocytosis of the viral vector.
NMR studies have confirmed the dicyclic conformation of the binding sequence, where cysteine bridges were formed between Cys11-Cys19, and Cys13-Cys31 [113].
Particularly the cyclic U11 sequence was able to competitively bind to uPAR at low dissociation constants within the nanomolar range [116-118]. It is clear from these findings that a loop-shaped structure of U-11 binds to the uPAR with high affinity and can therefore potentially be a promising targeting agent for cancer cells overexpressing the uPA receptor.
Previous investigators in the field of U11 binding have found a ligand affinity of Kd= 1.3-1.4μM by a displacement assay using the linear G-U11-G peptide (U11 peptide flanked at either terminus with Gly amino acid) as the displacing ligand against a fluorescently labelled uPA (uPA-FITC) (Figure 1-18). This calculated dissociation constant was found to be comparable to that of the linear RGD sequence (Kd= 1.0μM), although is also ten-fold lower than the dissociation constant of the cyclic RGD structure, indicating that the preservation of the natural conformation of the ligand is important in maintaining high receptor affinity. However, the targeting effect of the ligand is also dependent on characteristics of the nanoparticle onto which it is loaded, and not only on the binding affinity of the ligand itself. The process of sulphide cyclisation can be difficult, particularly if a third cysteine is required for conjugation onto the liposome.
61 0 20 40 60 80 100 120
1.E-10 1.E-09 1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 Gu11G Concentration (M) C e ll -a s s o c ia te d f lu o re s c e n c e ( % c o n tr o l)
Figure 1-18: Displacement of fluorescently labelled uPA by G-U11-G, found by flow cytometry assessment of U937 Human leukaemia cell line incubated with uPA-FITC in increasing concentrations of G-U11-G uPAR specific targeting peptide (from PhD thesis of Tom Kean, Cardiff Univeristy, 2006).
Having considered the range of information regarding uPAR-uPA interactions, the following highlights the advantages of U11 as a targeting ligand for uPAR-targeting:
a) uPAR is highly over-expressed in many types of cancer cells.
b) The U11 peptide is a relatively short sequence, allowing ease of synthesis and handling.
c) The peptide sequence contains aromatic amino acids (Trp, Phe and Tyr), which act as fluorescent probes for further characterisation.
d) U11 binds to the uPA-binding region of uPAR, therefore not blocking domain III of the receptor, the region where binding to α2MR occurs.